The filling of a hollow lumen within an intact glandular structure is a hallmark of early tumorigenesis. However the mechanisms involved in the formation of a glandular lumen and its filling during the development of carcinoma-in-situ are poorly understood. We have developed an in vitro three-dimensional (3D) culture model to investigate alterations in the architecture and growth properties of glandular epithelial structures (acini) during early stages of mammary carcinogenesis. We have found that formation of the lumen involves selective death of cells in the center of the acini. Filling of the luminal space, as provoked by oncogenes, requires not only induction of constitutive proliferation but also anti-apoptotic signals that allow cells to survive in the luminal space. We have identified two proapoptotic proteins that are involved in mediating cell death in this model: Bim, a proapoptotic BH3-only Bcl-2 family protein, and TRAIL, a death ligand. Bim was also found to be induced and critically involved in apoptosis induced by detachment of monolayer cells from matrix (a process referred to as anoikis) and several lines of evidence suggest that matrix deprivation may be involved in luminal cell death during morphogenesis. This proposal describes plans to elucidate the mechanisms that regulate cell death induced by Bim and TRAIL during both lumen formation and anoikis, identify other proteins that collaborate with Bim to mediate cell death, and determine the mechanisms whereby oncogenes allow cells to escape cell death induced by these apoptotic mediators. This study should provide important information relating to events associated with the development of carcinoma-in-situ tumors in humans. In addition, an understanding of the molecular events and cellular pathways utilized by oncogenes to allow survival in the luminal space may ultimately be valuable in the design of therapeutic strategies because targeting such pathways may promote selective death of tumor cells. Proapoptotic Bcl-2-family proteins are comprised of two subfamilies: BH3-only and multidomain Bax-like proteins. The studies in the first Aim will identify proteins from these two classes of proapoptotic Bcl-2 family proteins that are involved in luminal apoptosis and anoikis, define their role of in these death processes, and trace the pathways leading to their activation or inhibition. In addition, we will examine how oncogenes also escape from death processes regulated by these proteins. TRAIL is death-ligand family member that is induced during acinar morphogenesis in 3D cultures and its induction correlates temporally with the appearance of apoptotic cells in the future luminal space. Abrogation of TRAIL signaling has no effect on initial lumen formation but inhibits apoptosis of cells that proliferate into the luminal space after the lumen has formed. In Aim II, we will examine whether cells expressing proliferative oncogenes like cyclin D, HPVE7, and Wntl, that are unable to survive in the lumen, are protected from luminal cell death if TRAIL pathways are blocked by expression of truncated dominant inhibitory TRAIL receptors. In addition we will investigate the basis for the differential sensitivity of matrix-attached cells to TRAIL, and whether oncogenes that allow cells to fill the lumen escape TRAIL-induced cell death.